Hello everyone! We are dealing with a quite straightforward microarray experimental design on TIGR potato cDNA microarray. We Here is a brief description of the experiments: we hybridise virus- infected versus mock-infected potato plant RNA from several potato cultivars - each microarray is hybridised with RNA of the same cultivar - but there is no common reference (at least three biological replicates including dye swaps for each cultivar: balanced block design). Scanning and image analysis was done with ArrayPro Analyzer (Media Cybernetics). We also performed initial quality control and spot filtering in that software to remove "bad spots". Although we are not experienced R users we are trying to carry on the further analysis with limma package. And here are some of the problems/questions regarding limma: 1. As advised we would like to reduce the number of the spots before differential expression analysis (limma) - filter out genes. Does anyone have any suggestions which criteria should be used for gene filtering after normalisation (we have already removed "bad quality spots")? 2. Can be multifactorial testing done in limma, or do you suggest any other package (we would like to compare response to viral infection in 4 cultivars)? Does anyone have any suggestions on the design and contrast matrices? Thanks for any answers/comments, Matjaz Hren and Spela Baebler _____ National Institute of Biology Dept. of Plant Physiology and Biotechnology Ljubljana SLOVENIA _____

Hello everyone! We are struggling to implement R for the analysis of two color cDNA microarrays. Each microarray is hybridised with treated control plant RNA. There are 5 biological repetitions, they also include dye swaps. We would like to compare 4 groups of plants (cultivars) having the same treatment, meaning altogether the experiment has 20 microarrays. Initial quality control and spot filtering are performed in image analysis program. We are trying to do further analysis with limma package, although we are not experienced R users. We are wondering is limma the appropriate package for such an experimental outline, or we should try any other packages? Which ones would you recommend? Thanks for any answers/comments, Spela Baebler National Institute of Biology Dept. of Plant Physiology and Biotechnology Ljubljana SLOVENIA _____

Nice to get a reply:-) We now have a problem which is more statistics than R. We have a problem with creating a design and contrast matrix. If we just want to compare vithin one variety, it is easy design <- c(1,-1,1,-1,1) where we just take into the account the dye swaps. We couldn't find any example that would fit our experimental design in Limma users Guide, since we don't have a common reference for all arrays. So do you have a suggestion for contrast matrix if we want to compare the 4 varieties? Waiting for answers... Spela P.S:: Is there someone using TIGR potato microarrays out there? -----Original Message----- From: Gordon K Smyth [mailto:smyth@wehi.EDU.AU] Sent: Thursday, July 21, 2005 12:58 AM To: ?pela Baebler Cc: bioconductor at stat.math.ethz.ch Subject: [BioC] Gene filtering for differential expression (limma) > Date: Tue, 19 Jul 2005 10:53:37 +0200 > From: ?pela Baebler <spela.baebler at="" nib.si=""> > Subject: [BioC] Gene filtering for differential expression (limma) > To: <bioconductor at="" stat.math.ethz.ch=""> > > Hello everyone! > > We are struggling to implement R for the analysis of two color cDNA microarrays. > > Each microarray is hybridised with treated control plant RNA. There are 5 biological repetitions, > they also include dye swaps. We would like to compare 4 groups of plants (cultivars) having the > same treatment, meaning altogether the experiment has 20 microarrays. Initial quality control and > spot filtering are performed in image analysis program. > > We are trying to do further analysis with limma package, although we are not experienced R users. > > We are wondering is limma the appropriate package for such an experimental outline, or we should > try any other packages? Which ones would you recommend? limma is certainly intended to address this sort of experiment. Have you run into problems or are you just trying to get an idea of the alternatives? Gordon > Thanks for any answers/comments, > > Spela Baebler > > National Institute of Biology > Dept. of Plant Physiology and Biotechnology > Ljubljana > SLOVENIA

Gordon's right, if you post an example of your targets file we will have a better idea of what you mean >>Initial quality control and >>spot filtering are performed in image analysis program. Personally, I wouldn't recommend doing this. The way R works, it's better to have all your data points present in all files. I leave all data in and flag up bad spots, and remove them at the end of the analysis, not the beginning :-) -----Original Message----- From: bioconductor-bounces@stat.math.ethz.ch [mailto:bioconductor-bounces at stat.math.ethz.ch] On Behalf Of Gordon Smyth Sent: 21 July 2005 10:30 To: Spela Baebler Cc: bioconductor at stat.math.ethz.ch Subject: Re: [BioC] Gene filtering for differential expression (limma) At 06:10 PM 21/07/2005, =?iso-8859-2?Q?=A9pela_Baebler?= wrote: >Nice to get a reply:-) >We now have a problem which is more statistics than R. We have a >problem >with creating a design and contrast matrix. > >If we just want to compare vithin one variety, it is easy design <- >c(1,-1,1,-1,1) where we just take into the account the dye swaps. > >We couldn't find any example that would fit our experimental design in >Limma users Guide, since we don't have a common reference for all arrays. >So do you have a suggestion for contrast matrix if we want to compare the >4 varieties? Perhaps you need to read the documentation more carefully because there are examples in the User's Guide of two colour experiments without a common reference. For example, Section 9.4 "Direct Designs" considers an example with 3 treatments. Do you mean "design matrix" rather than "contrast matrix", or are you implying you already have a design matrix? Have you tried reading in your targets file and using modelMatrix()? >Waiting for answers... To make good use of this mailing list you need to read the documentation carefully, get as far as you can on your own, and then ask questions which are specific as possible. Just waiting for complete answers is not on. BTW, what does your question have to do with "gene filtering"? Gordon >Spela > >P.S:: Is there someone using TIGR potato microarrays out there? > >-----Original Message----- >From: Gordon K Smyth [mailto:smyth at wehi.EDU.AU] >Sent: Thursday, July 21, 2005 12:58 AM >To: (c)pela Baebler >Cc: bioconductor at stat.math.ethz.ch >Subject: [BioC] Gene filtering for differential expression (limma) > > > Date: Tue, 19 Jul 2005 10:53:37 +0200 > > From: ?pela Baebler <spela.baebler at="" nib.si=""> > > Subject: [BioC] Gene filtering for differential expression (limma) > > To: <bioconductor at="" stat.math.ethz.ch=""> > > > > Hello everyone! > > > > We are struggling to implement R for the analysis of two color cDNA > microarrays. > > > > Each microarray is hybridised with treated control plant RNA. There > > are > 5 biological repetitions, > > they also include dye swaps. We would like to compare 4 groups of > plants (cultivars) having the > > same treatment, meaning altogether the experiment has 20 > microarrays. Initial quality control and > > spot filtering are performed in image analysis program. > > > > We are trying to do further analysis with limma package, although we > are not experienced R users. > > > > We are wondering is limma the appropriate package for such an > experimental outline, or we should > > try any other packages? Which ones would you recommend? > >limma is certainly intended to address this sort of experiment. Have >you >run into problems or are >you just trying to get an idea of the alternatives? > >Gordon > > > Thanks for any answers/comments, > > > > Spela Baebler > > > > National Institute of Biology > > Dept. of Plant Physiology and Biotechnology > > Ljubljana > > SLOVENIA _______________________________________________ Bioconductor mailing list Bioconductor at stat.math.ethz.ch https://stat.ethz.ch/mailman/listinfo/bioconductor